ElevenLabs Music v1: FAQ

ElevenLabs Music v1 composes full musical tracks from a text description, with control over genre, mood, tempo, instruments, and key.

1. What is ElevenLabs Music v1?

It is a text-to-music model. You describe the music you want in plain language and it composes a complete track, with or without vocals. It is built for songs and musical beds, not short spot effects.

2. How do I prompt it?

Write a clean description of the music you want, including genre, mood, tempo, and instruments. The prompt accepts up to 7,000 characters, so you have room for detail, but keep it focused on the sound rather than long instructions. You can add precise control cues such as "120 BPM" or "in A minor".

3. How long can a track be?

Duration is set in seconds, from a minimum of 10 seconds up to a maximum of 300 seconds (5 minutes). The default is 30 seconds.

4. Can it make instrumentals and vocals?

Yes to both. For an instrumental, include "instrumental only" in the prompt. For vocals, specify the language or style, and you can supply sample lyric lines. You can also direct timing, for example "lyrics begin at 15 seconds".

5. Can I reference a specific artist?

No. Real artist, composer, or musician names are blocked and will cause the generation to fail. Describe the intent through style instead: use genre, mood, instrument, and tempo descriptors (for example "epic cinematic orchestral" rather than a composer's name).

6. What is it not for?

Short sound effects. For explosions, footsteps, UI sounds, and other brief foley, use the sound effects model instead.

At a glance